| Food texture is the main index in evaluating food quality.There are two methods to evaluate food texture,including sensory evaluation method and instrument measuring method.The former method obtains texture features by mastication early and later which is attachs to the human physiological perception,so the evaluation results are easily affected by subjective factors,therefore,this method is lack of accuracy and stability.The latter obtains texture features by cylindrical or wedge-shape indenters of texture analyzer to extrude foods to obtain stress strain curves,which is different from the surface morphology of human teeth,and texture analyzer is difficult to realize some texture features by multiple compressions,such as chewiness.Furthermore,a various kinds of food boluses are obtained by human mastication,and masticatory performance reflects the ability of human broken food and is applied to analyze food boluses.At present,masticatory performance evaluation methods can not effectively analyze different textures of food boluses.Therefore,it is necessary to find reasonable evaluation methods of food texture and masticatory performance to provide a new method for food texture evaluation and technology support for designing the foods which are suitable for the patients who have oral diseases to increase the masticatory performance and swallowing security.According to defects of previous methods,masticatory performance method was optimized to analyze food boluses properties.The relationship between the electromyographic signals in the process of chewing and food texture was researched.Double-teeth indenter was designed to evaluate food texture and simulate broken food process.Then simulated masticatory system was established.The purpose of this paper is to develop the masticating simulation that can evaluate food texture and to obtain the bolus close to human bolus properties.The main research contents in this paper are as follows:(1)Methods of evaluating masticatory performance for two kinds of food bolus were determined and meanwhile particle size distribution and image texture properties of food bolus were researched.For the food boluses which were composed of independent particles and saliva mixed after mastication,we use the median particle size(d50)obtained by sieving weight method to evaluate the masticatory performance.Six kinds of foods which were peanut,carrot,apple,pickle,bean curd and ham sausage were masticated by subjects for 5,10 cycles and swallow cycles and the boluses were collected and tested by sieving weight method.The results showed that recovered mass of food bolus at swallow cycles were lower than recovered mass at 5,10 cycles.d50 values were significantly different among different cycles and different food samples(p<0.05).For the cohesive bolus after mastication,such as bread,image processing method was used to evaluate the masticatory performance.Gray-level gradient co-occurrence matrix(GGCM)and gray-level co-occurrence matrix(GLCM)were used to extract the image texture features.Stepwise discriminant analysis results showed that the accuracies that the images were correctly classified into their corresponding chewing cycles by GGCM and GLCM were 75.30%and 71.95%,respectively.Inverse different moment(IDM)was selected as the main parameter to describe the decomposition of boluses by principal component analysis.Pearson correlation coefficients showed that the correlation between IDM of GGCM(IDMGGCM)and weigh gains,texture properties were better than the correlation between IDM of GLCM(IDMGLCM)and weigh gains,texture properties.Therefore,GGCM was used to extract bread bolus images and IDMGGCM was able to effectively reflect the image texture properties.(2)Relationship between the electromyographic signals during chewing process and food texture was studied.Various kinds of texture foods were masticated by ten subjects.The chewing processes were recorded by electromyography,and food textures were measured by sensory assessment and universal instrument measurement.Experiment results showed that the electromyographic signals of the seven foods were significantly different,first 3 chews-EMG activities per cycles(F3-wc)were significantly different,middle 3 chews-EMG activities per cycles(M3-wc)and last 3 chews-EMG activities per cycles(L3-wc)were not different among seven foods.Pearson correlation coefficients showed that the correlation was positive between hardness values which were tested by sensory assessment and universal instrument measurement and F3-wc(correlation coefficient were 0.924 and 0.872,p<0.01,respectively),and no correlation was observed between hardness values which were tested by sensory assessment and universal instrument measurement and M3-wc or L3-wc.The correlation between electromyographic signals and sensory assessment for every food sample was better than that between universal instrument measurement and electromyographic signals.(3)Food texture assessment and crushing effect for artificial indenters were studied.Four artificial indenters which were double-teeth indenter,single-teeth indenter,cylinder-type indenter and wave-type indenter were developed by the technology of bionic and precision machining,then assembled on texture analyzer.They were used to measure food texture and crushing effect,then the correlation between food texture with electromyographic signals was studied.The results showed that Pearson correlation coefficients between texture properties by double-teeth indenter and sensory assessment were better than other indenters.At deformation rate 90%,the median particle size dso(r=0.921,p<0.01)and image texture IDMGGCM(r=0.851,p<0.01)of food bolus with double-teeth indenter were closest to those obtained from human.Single-teeth indenter,cylinder-type indenter and wave-type indenter acquired more cycles to equal to food bolus properties obtained from human.Double-teeth indenter was better than the other indenters in food texture assessment and crushing effect.(4)Crushing efficiency for artificial indenters was studied based on the finite element method.Three-dimensional models of double-teeth indenter,single-teeth indenter,cylinder-type indenter and wave-type indenter were established by reverse engineering.Masticatory motions of molars could be simplified into two kinds of modes which were compressing mode and grinding mode.Takeing linear elastic and hyperelastic materials for example,two modes and two kinds of materials of masticatory motions for four indenters were established,stress distribution on surface of material after chewing was studied by joint strength theory.The results showed that the stress values on linear elastic material for four indenters were higher than those on hyperelastic material.Mises stress,shear stress and primary stress for double-teeth indenter in compressing mode and grinding mode were higher than single-teeth indenter,cylinder-type indenter and wave-type indenter.(5)Food texture assessment and crushing effect for masticating simulation were researched under multiple-time chewing mode.Test materials were peanut,carrot,apple,pickle,bean curd,ham sausage and bread.Seven kinds of foods were chewed by masticating simulation to evaluate food texture.It was showed that masticating simulation can evaluate food textures(p<0.01)and coefficients of correlation between measurement of masticating simulation and sensory assessment was better than universal instrument measurement and sensory assessment to every food sample.Coefficients of correlation between hardness measured by masticating simulation and electromyographic signals(r=0.922,p<0.01)were more close to coefficients of correlation between hardness measured by sensory assessment and electromyographic signals(r=0.924,p<0.01).The median particle size(dso)and image texture(IDMGGCM)of food bolus with masticating simulation were closer to food bolus properties obtained from human,recovered mass at swallow cycles reached 50%.Masticating simulation for evaluating food texture was better than universal instrument measurement,and the particle properties and the image texture properties of food boluses produced by masticating simulation is able to represent those of food boluses produced by human.In conclusion,research of texture evaluation and food bolus properties based on masticating simulation can present a new method for food texture evaluation and it provides a technology support for designing foods that are suitable for oral disease patients.Meanwhile,this research can supply theoretical foundation and important basis for mastication mechanism. |
PDF Full Text Request